Providers of dynamic Web applications are currently unable to accommodate heavy usage without significant investment in infrastructure and in-house management capability. Our goal is to develop technology to enable a third party to offer scalability as a subscription service with "per-click" pricing to application providers. To this end we are building a proxy caching system that will scale delivery of dynamic Web content to a large number of users. We have a working prototype, which is presently under heavy construction and looks like this (conceptually):

From a research standpoint, the key challenges include:

• Creating fully distributed mechanisms to manage consistency of cached data

• Developing flexible consistency models that permit bending of consistency rules as needed to avoid overloading application servers

• Studying "load-aware" cache management policies that never place excessive load on application servers

• Developing means of ensuring data security and privacy while still permitting efficient and effective consistency management

Publications

A Scalability Service for Dynamic Web Applications. Christopher Olston, Amit Manjhi, Charles Garrod, Anastassia Ailamaki, Bruce M. Maggs, and Todd C. Mowry. 2nd Biannual Conference on Innovative Data Systems Research (CIDR 2005).

Simultaneous Scalability and Security for Data-Intensive Web Applications. Amit Manjhi, Christopher Olston, Anastassia Ailamaki, Bruce M. Maggs, Todd C. Mowry, and Anthony Tomasic. 25th ACM International Conference on the Managament of Data (SIGMOD 2006).

People

Faculty and collaborators:

• Anastassia Ailamaki

• Michael Kozuch (Intel Research Pittsburgh)

• Phillip B. Gibbons

• Bruce Maggs

• Todd Mowry

• Christopher Olston

• Dawn Song

• Anthony Tomasic

• Haifeng Yu (Intel Research Pittsburgh)

Students:

• Charles Garrod

• Amit Manjhi